Clinical characterization of FGFR2b expression in patients with advanced gastric or gastroesophageal junction adenocarcinoma
Seiya Sato, Shannon Rhodes, Yu Aoki, Izuma Nakayama, Tadayoshi Hashimoto, Joseph Hawkins, Rolando E. Yanes, Chia‐Hsuin Chang, Yuichi Nakamura, Aki Kawazoe, Saori Mishima, Daisuke Kotani, Yoshinori Kuboki, H. Bando, Takashi Kojima, S. Yoshino, Stephanie Reading, Hayden Honeycutt, Elizabeth Finger, Itaru Endo, Naoya Sakamoto, Taiji Kuwata, T. Yoshino, Kohei Shitara
Abstract
BACKGROUND: Fibroblast growth factor receptor 2b (FGFR2b) is a novel protein biomarker expressed in gastric and gastroesophageal junction tumors (GC/GEJC). Phase III trials are evaluating the efficacy and safety of FGFR2b-targeting therapies. However, features of FGFR2b-expressing tumors and co-occurrence of FGFR2b expression with currently actionable biomarkers in gastric cancer remain unclear. MATERIALS AND METHODS: We carried out a single-institution retrospective cohort study of patients who initiated systemic therapy for GC/GEJC to evaluate features of FGFR2b-positive tumors, first-line (1L) treatments received, and co-occurrence with select actionable biomarkers [human epidermal growth factor receptor 2 (HER2), mismatch repair, programmed death-ligand 1 (PD-L1), claudin-18 isoform 2]. A sample was deemed FGFR2b-positive when any (>0%) tumor cells exhibiting moderate (2+) or strong (3+) membrane staining were detected ('any 2+/3+'). Other biomarkers were assessed based on current clinical guidelines. FGFR2b-stratified real-world overall survival (OS) was estimated using the Kaplan-Meier and Cox proportional regression models. RESULTS: Of 547 GC/GEJC patients identified, 492 (89.9%) met inclusion/exclusion criteria, had evaluable FGFR2b staining, and had complete clinical data. Estimated prevalence of FGFR2b any 2+/3+ was 15.4% [95% confidence interval (CI) 12.4% to 19.0%] in the full cohort of patients and 29.8% (95% CI 22.0% to 38.7%) in patients with samples collected within 1.5 years of study initiation. The majority (53.9%; 95% CI 42.1% to 65.5%) of FGFR2b any 2+/3+ tumor specimens were negative for other assessed biomarkers at a PD-L1 cut-off of combined positive score ≥5. In HER2-negative patients treated in 1L with chemotherapy alone, median OS was 11.5 months (95% CI 10.0-16.3 months) and 15.3 months (95% CI 13.0-16.8 months) for FGFR2b any 2+/3+ and FGFR2b 0/1+, respectively. There was no association between FGFR2b overexpression level and OS [adjusted hazard ratio (HR) 1.14, 95% CI 0.84-1.55]. CONCLUSIONS: This study revealed limited overlap of FGFR2b overexpression with currently actionable biomarkers, suggesting FGFR2b is a novel biomarker that identifies a distinct GC/GEJC patient population who may benefit most from an FGFR2b-targeting therapy.